mirrors/docling
1
0
Fork 0
mirror of https://github.com/DS4SD/docling.git synced 2025-07-30 22:14:37 +00:00
Code Issues Packages Projects Releases Wiki Activity
baaeb60b4a
docling/tests/data/docx/unit_test_lists.docx.json
Peter Staar baaeb60b4a add real e2e tests for html and docx 
Signed-off-by: Peter Staar <taa@zurich.ibm.com>
2024-10-29 05:36:23 +01:00

921 lines
17 KiB
JSON
Raw Blame History

{
"schema_name": "DoclingDocument",
"version": "1.0.0",
"name": "unit_test_lists",
"origin": {
"mimetype": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
"binary_hash": 13601004233111293776,
"filename": "unit_test_lists.docx"
},
"furniture": {
"self_ref": "#/furniture",
"children": [],
"name": "_root_",
"label": "unspecified"
},
"body": {
"self_ref": "#/body",
"children": [
{
"$ref": "#/groups/0"
}
],
"name": "_root_",
"label": "unspecified"
},
"groups": [
{
"self_ref": "#/groups/0",
"parent": {
"$ref": "#/body"
},
"children": [
{
"$ref": "#/texts/0"
}
],
"name": "header-0",
"label": "section"
},
{
"self_ref": "#/groups/1",
"parent": {
"$ref": "#/texts/0"
},
"children": [
{
"$ref": "#/texts/7"
},
{
"$ref": "#/texts/12"
},
{
"$ref": "#/texts/17"
},
{
"$ref": "#/texts/25"
},
{
"$ref": "#/texts/30"
},
{
"$ref": "#/texts/36"
}
],
"name": "header-2",
"label": "section"
},
{
"self_ref": "#/groups/2",
"parent": {
"$ref": "#/texts/7"
},
"children": [
{
"$ref": "#/texts/8"
},
{
"$ref": "#/texts/9"
},
{
"$ref": "#/texts/10"
}
],
"name": "list",
"label": "list"
},
{
"self_ref": "#/groups/3",
"parent": {
"$ref": "#/texts/12"
},
"children": [
{
"$ref": "#/texts/13"
},
{
"$ref": "#/texts/14"
},
{
"$ref": "#/texts/15"
}
],
"name": "list",
"label": "list"
},
{
"self_ref": "#/groups/4",
"parent": {
"$ref": "#/texts/17"
},
"children": [
{
"$ref": "#/texts/18"
},
{
"$ref": "#/texts/19"
},
{
"$ref": "#/groups/5"
},
{
"$ref": "#/texts/23"
}
],
"name": "list",
"label": "list"
},
{
"self_ref": "#/groups/5",
"parent": {
"$ref": "#/groups/4"
},
"children": [
{
"$ref": "#/texts/20"
},
{
"$ref": "#/texts/21"
},
{
"$ref": "#/texts/22"
}
],
"name": "list",
"label": "list"
},
{
"self_ref": "#/groups/6",
"parent": {
"$ref": "#/texts/25"
},
"children": [
{
"$ref": "#/texts/26"
},
{
"$ref": "#/groups/7"
},
{
"$ref": "#/texts/28"
}
],
"name": "list",
"label": "list"
},
{
"self_ref": "#/groups/7",
"parent": {
"$ref": "#/groups/6"
},
"children": [
{
"$ref": "#/texts/27"
}
],
"name": "list",
"label": "list"
},
{
"self_ref": "#/groups/8",
"parent": {
"$ref": "#/texts/30"
},
"children": [
{
"$ref": "#/texts/31"
},
{
"$ref": "#/groups/9"
},
{
"$ref": "#/texts/34"
}
],
"name": "list",
"label": "list"
},
{
"self_ref": "#/groups/9",
"parent": {
"$ref": "#/groups/8"
},
"children": [
{
"$ref": "#/texts/32"
},
{
"$ref": "#/groups/10"
}
],
"name": "list",
"label": "list"
},
{
"self_ref": "#/groups/10",
"parent": {
"$ref": "#/groups/9"
},
"children": [
{
"$ref": "#/texts/33"
}
],
"name": "list",
"label": "list"
},
{
"self_ref": "#/groups/11",
"parent": {
"$ref": "#/texts/36"
},
"children": [
{
"$ref": "#/texts/37"
},
{
"$ref": "#/texts/38"
},
{
"$ref": "#/groups/12"
},
{
"$ref": "#/texts/42"
}
],
"name": "list",
"label": "list"
},
{
"self_ref": "#/groups/12",
"parent": {
"$ref": "#/groups/11"
},
"children": [
{
"$ref": "#/texts/39"
},
{
"$ref": "#/texts/40"
},
{
"$ref": "#/groups/13"
}
],
"name": "list",
"label": "list"
},
{
"self_ref": "#/groups/13",
"parent": {
"$ref": "#/groups/12"
},
"children": [
{
"$ref": "#/texts/41"
}
],
"name": "list",
"label": "list"
}
],
"texts": [
{
"self_ref": "#/texts/0",
"parent": {
"$ref": "#/groups/0"
},
"children": [
{
"$ref": "#/texts/1"
},
{
"$ref": "#/texts/2"
},
{
"$ref": "#/texts/3"
},
{
"$ref": "#/texts/4"
},
{
"$ref": "#/texts/5"
},
{
"$ref": "#/texts/6"
},
{
"$ref": "#/groups/1"
}
],
"label": "section_header",
"prov": [],
"orig": "Test Document",
"text": "Test Document",
"level": 1
},
{
"self_ref": "#/texts/1",
"parent": {
"$ref": "#/texts/0"
},
"children": [],
"label": "paragraph",
"prov": [],
"orig": "",
"text": ""
},
{
"self_ref": "#/texts/2",
"parent": {
"$ref": "#/texts/0"
},
"children": [],
"label": "paragraph",
"prov": [],
"orig": "",
"text": ""
},
{
"self_ref": "#/texts/3",
"parent": {
"$ref": "#/texts/0"
},
"children": [],
"label": "paragraph",
"prov": [],
"orig": "Paragraph 2.1.1",
"text": "Paragraph 2.1.1"
},
{
"self_ref": "#/texts/4",
"parent": {
"$ref": "#/texts/0"
},
"children": [],
"label": "paragraph",
"prov": [],
"orig": "",
"text": ""
},
{
"self_ref": "#/texts/5",
"parent": {
"$ref": "#/texts/0"
},
"children": [],
"label": "paragraph",
"prov": [],
"orig": "Paragraph 2.1.2",
"text": "Paragraph 2.1.2"
},
{
"self_ref": "#/texts/6",
"parent": {
"$ref": "#/texts/0"
},
"children": [],
"label": "paragraph",
"prov": [],
"orig": "",
"text": ""
},
{
"self_ref": "#/texts/7",
"parent": {
"$ref": "#/groups/1"
},
"children": [
{
"$ref": "#/groups/2"
},
{
"$ref": "#/texts/11"
}
],
"label": "section_header",
"prov": [],
"orig": "Test 1:",
"text": "Test 1:",
"level": 3
},
{
"self_ref": "#/texts/8",
"parent": {
"$ref": "#/groups/2"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 1",
"text": "List item 1",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/9",
"parent": {
"$ref": "#/groups/2"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 2",
"text": "List item 2",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/10",
"parent": {
"$ref": "#/groups/2"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 3",
"text": "List item 3",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/11",
"parent": {
"$ref": "#/texts/7"
},
"children": [],
"label": "paragraph",
"prov": [],
"orig": "",
"text": ""
},
{
"self_ref": "#/texts/12",
"parent": {
"$ref": "#/groups/1"
},
"children": [
{
"$ref": "#/groups/3"
},
{
"$ref": "#/texts/16"
}
],
"label": "section_header",
"prov": [],
"orig": "Test 2:",
"text": "Test 2:",
"level": 3
},
{
"self_ref": "#/texts/13",
"parent": {
"$ref": "#/groups/3"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item a",
"text": "List item a",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/14",
"parent": {
"$ref": "#/groups/3"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item b",
"text": "List item b",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/15",
"parent": {
"$ref": "#/groups/3"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item c",
"text": "List item c",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/16",
"parent": {
"$ref": "#/texts/12"
},
"children": [],
"label": "paragraph",
"prov": [],
"orig": "",
"text": ""
},
{
"self_ref": "#/texts/17",
"parent": {
"$ref": "#/groups/1"
},
"children": [
{
"$ref": "#/groups/4"
},
{
"$ref": "#/texts/24"
}
],
"label": "section_header",
"prov": [],
"orig": "Test 3:",
"text": "Test 3:",
"level": 3
},
{
"self_ref": "#/texts/18",
"parent": {
"$ref": "#/groups/4"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 1",
"text": "List item 1",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/19",
"parent": {
"$ref": "#/groups/4"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 2",
"text": "List item 2",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/20",
"parent": {
"$ref": "#/groups/5"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 1.1",
"text": "List item 1.1",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/21",
"parent": {
"$ref": "#/groups/5"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 1.2",
"text": "List item 1.2",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/22",
"parent": {
"$ref": "#/groups/5"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 1.3",
"text": "List item 1.3",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/23",
"parent": {
"$ref": "#/groups/4"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 3",
"text": "List item 3",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/24",
"parent": {
"$ref": "#/texts/17"
},
"children": [],
"label": "paragraph",
"prov": [],
"orig": "",
"text": ""
},
{
"self_ref": "#/texts/25",
"parent": {
"$ref": "#/groups/1"
},
"children": [
{
"$ref": "#/groups/6"
},
{
"$ref": "#/texts/29"
}
],
"label": "section_header",
"prov": [],
"orig": "Test 4:",
"text": "Test 4:",
"level": 3
},
{
"self_ref": "#/texts/26",
"parent": {
"$ref": "#/groups/6"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 1",
"text": "List item 1",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/27",
"parent": {
"$ref": "#/groups/7"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 1.1",
"text": "List item 1.1",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/28",
"parent": {
"$ref": "#/groups/6"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 2",
"text": "List item 2",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/29",
"parent": {
"$ref": "#/texts/25"
},
"children": [],
"label": "paragraph",
"prov": [],
"orig": "",
"text": ""
},
{
"self_ref": "#/texts/30",
"parent": {
"$ref": "#/groups/1"
},
"children": [
{
"$ref": "#/groups/8"
},
{
"$ref": "#/texts/35"
}
],
"label": "section_header",
"prov": [],
"orig": "Test 5:",
"text": "Test 5:",
"level": 3
},
{
"self_ref": "#/texts/31",
"parent": {
"$ref": "#/groups/8"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 1",
"text": "List item 1",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/32",
"parent": {
"$ref": "#/groups/9"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 1.1",
"text": "List item 1.1",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/33",
"parent": {
"$ref": "#/groups/10"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 1.1.1",
"text": "List item 1.1.1",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/34",
"parent": {
"$ref": "#/groups/8"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 3",
"text": "List item 3",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/35",
"parent": {
"$ref": "#/texts/30"
},
"children": [],
"label": "paragraph",
"prov": [],
"orig": "",
"text": ""
},
{
"self_ref": "#/texts/36",
"parent": {
"$ref": "#/groups/1"
},
"children": [
{
"$ref": "#/groups/11"
},
{
"$ref": "#/texts/43"
},
{
"$ref": "#/texts/44"
},
{
"$ref": "#/texts/45"
}
],
"label": "section_header",
"prov": [],
"orig": "Test 6:",
"text": "Test 6:",
"level": 3
},
{
"self_ref": "#/texts/37",
"parent": {
"$ref": "#/groups/11"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 1",
"text": "List item 1",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/38",
"parent": {
"$ref": "#/groups/11"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 2",
"text": "List item 2",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/39",
"parent": {
"$ref": "#/groups/12"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 1.1",
"text": "List item 1.1",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/40",
"parent": {
"$ref": "#/groups/12"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 1.2",
"text": "List item 1.2",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/41",
"parent": {
"$ref": "#/groups/13"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 1.2.1",
"text": "List item 1.2.1",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/42",
"parent": {
"$ref": "#/groups/11"
},
"children": [],
"label": "list_item",
"prov": [],
"orig": "List item 3",
"text": "List item 3",
"enumerated": false,
"marker": "-"
},
{
"self_ref": "#/texts/43",
"parent": {
"$ref": "#/texts/36"
},
"children": [],
"label": "paragraph",
"prov": [],
"orig": "",
"text": ""
},
{
"self_ref": "#/texts/44",
"parent": {
"$ref": "#/texts/36"
},
"children": [],
"label": "paragraph",
"prov": [],
"orig": "",
"text": ""
},
{
"self_ref": "#/texts/45",
"parent": {
"$ref": "#/texts/36"
},
"children": [],
"label": "paragraph",
"prov": [],
"orig": "",
"text": ""
}
],
"pictures": [],
"tables": [],
"key_value_items": [],
"pages": {}
}
Reference in New Issue View Git Blame Copy Permalink